AI accelerates business formation and productivity across the economy
Updated
Updated · Apollo Global Management · May 5
AI accelerates business formation and productivity across the economy
8 articles · Updated · Apollo Global Management · May 5
The report says the disruption is hitting cognitive and white-collar jobs, echoing the post-2001 China shock while keeping overall US unemployment relatively low through offsetting gains.
It argues AI is already causing immediate losses in exposed sectors but also helping businesses expand, as cheaper Chinese imports once did, supporting hiring and new economic activity.
Citing research from the IMF, ECB, BIS and others, it says half of post-1980 job growth came from occupations that did not exist in 1980.
The 'AI shock' is compared to past industrial shifts, but what if this time is truly different for white-collar workers?
AI is creating new jobs, but will they offer the same career stability and upward mobility as the roles they are replacing?
As AI boosts the digital economy, can our physical infrastructure, like the power grid, actually keep up with its massive demands?
The AI Productivity Paradox: Minimal Gains for Most, Massive Value for 20% of Leaders
Overview
By early 2026, widespread AI adoption across major economies has led to minimal productivity gains for most firms, mainly due to superficial implementation, skill gaps, and governance issues like unauthorized AI use. However, about 20% of firms, identified as AI leaders, achieve significant value by strategically reinventing their business models, leveraging industry convergence, and building strong governance and trust. This approach triggers a powerful data flywheel effect, amplifying growth and competitive advantage. Meanwhile, small businesses are rapidly adopting AI to lower barriers and boost efficiency, though they face risks that require responsible leadership. The future depends on broadening AI benefits through strategic scaling, workforce investment, and inclusive policies to avoid widening divides.